The overall goal is to assess the minimum fraction of childhood cancers which may be attributable to major genes through proband and family studies, to characterize those genes at the clinical or phenotypic level with respect to cancer risks, and to determine the impact of those genes on long term survivors of childhood cancer with respect to their risks of second tumors (including those second tumors apparently related to radiation and/or chemotherapy) and the risk of cancer in their offspring. Preliminary data suggest that sarcomas are the most frequent second tumor among childhood cancer survivors, and that they arise primarily, if not exclusively, in children with evidence of genetic predisposing factors such as bilateral or familial tumors. The immediate goal is to determine the frequency of sarcomas as second malignant tumors among an estimated 1350 3-year survivors of childhood cancer at risk, and to determine the role of genetic and treatment factors in this outcome as a model for studying genetic and environmental interaction in cancer.